The present invention relates generally to computer network-based messaging systems, including so-called chat rooms, and more specifically to a method of filtering undesirable words and phrases in a chat room and to a messaging system including a message filter incorporating such method.
With the evolution of increasingly sophisticated Internet tools and the widespread availability of broadband Internet connections, the World Wide Web (“Web”) experience is moving steadily beyond the passive dissemination of information, towards real-time interaction between simultaneous users. Virtual communities exist for groups that share every conceivable interest, hobby, or profession. Increasingly more people of all ages use the Internet as a place to meet and to interact with other people for work and for play. As a consequence, chat rooms have become ubiquitous on the Internet, and accordingly, the maintenance of behavioral standards and safety, especially for young people and minors, is becoming a major societal concern.
Consequently, administrators of chat sites face challenges related to maintaining standards of conduct to prevent a site from degenerating into a forum for types of discussion that were never intended. One such challenge is how to maintain standards within an environment like the Internet where the participants are anonymous and therefore cannot be held accountable with traditional methods. Around-the-clock real-time monitoring by the administrator is not economically feasible for most Internet businesses.
One common approach is to use basic word filters to eliminate offensive words and profanity from the chat conversation. Such filters, often referred to as “black lists” or “exclusive” filters, generally involve blocking words and/or phrases contained in a black list or exclusive dictionary. Unfortunately such simplistic exclusive dictionary approaches can never be exhaustive and are easily outwitted by the well-known practice of “obfuscation.” This practice, also known by other labels such as “creative misspelling,” involves users purposely misspelling words or phrases in a chat message in order to prevent the detection of the words or phrases by exclusive word filters, while still permitting the meaning of the words to be understood by human readers.
Another limitation of these exclusive filters is that, depending on the needs of a chat site, certain words and phrases that are neither profanity nor generally offensive may need to be discouraged in order to preserve certain specific site standards. For example, in a community site for children who do not fully grasp the importance of password safety, phrases like “what's your password,” “gimme your pass,” and “my password is” need to be discouraged. Furthermore, these needs arise dynamically out of the needs of a community and continually evolve. Accordingly, the ongoing task of maintaining an exclusive dictionary of all prohibited words and phrases requires devoting a tremendous amount of resources to constantly monitoring chat room conversations and updating the exclusive dictionary as new offensive language evolves. This is not practical or economically feasible for most Internet businesses. Furthermore, there will always be some delay in updating the exclusive dictionary based on monitoring, since it is nearly impossible for the administrator to anticipate all of the creative misspellings and word combinations that will need to be blocked.
Another approach is to use a more restrictive form of message filtering or chat filtering, known is “white list” or “inclusive” filtering. Such filters prevent the use of offensive words, as well as the use of word obfuscation, since they only allow words contained in a list of permitted words that have been preapproved for use in chatting. This list of permitted words in often referred to as a “white list” or “inclusive dictionary.” However, in order to effectively block the use of offensive language, the inclusive dictionary must be kept relatively small, which can significantly hinder the ability of users to effectively chat with each other. Further, users often adopt “workarounds” by using permitted words in a context that implies offensive or otherwise inappropriate messages.
As described in U.S. Patent Application Publication No. 2009/0228557 to Ganz et al., a hybrid or multiple-layer chat filter was developed which uses a combination of an exclusive dictionary and an inclusive dictionary. While the system described therein overcomes many of the disadvantages of using either method alone, as with other inclusive dictionary-based filters, users are still limited to the words provided in the inclusive diction when formulating a message.